Full Winter Forecast 2024-25

[psuhoffman]

Sorry this is a little later than usual, I've been busy with some new responsibilities at work.  I will also keep this generally brief as I think we all know the general ideas.  

Major factors
PDO and ENSO:  We are currently in a cold neutral enso state and expected to remain cold neutral or weak nina status.  We are also still entrenched in a deeply negative PDO regime, coming off one of the most negative 3 month periods on record.  However, the PDO does show signs if creeping back towards neutral although it is likely to remain solidly negative through winter based on historical analysis of where we are in the cycle. 

I don't think it really matters all that much if enso is cold neutral or weak nina, if it ends up -.3 or -.7 for the winter...when scouring all the analogs it really didn't make much difference.  The flavor of all those -PDO cold enso winters was pretty similar regardless of the exact -enso strength.  

There are two types of cold enso years...ones that feature a more flat pacific ridge and they are generally warm snowless winters.  Think 2002, 2008, 2012, and 2023.  Then there are cold enso seasons with a more poleward pacific ridge which tend to be colder across eastern north america.  Think 2006, 2009, 2011, 2014, 2018, 2022.  The colder years do not always mean significantly more snow though as they are still generally dry and we are stuck between weak southern sliders and northern stream systems in the flow.  But determining which flavor the enso will take will determine temps and the chance we have some snow and perhaps if lucky close to median snowfall.  

Fall Pattern: We were running a consistent -PNA driven pattern through early fall with a warm dry dome of high pressure over eastern N America.  We have since transitioned into a -EPO -NAO pattern that shows signs of persisting in some form into and perhaps through December.  

Minor factors:

QBO: We are currently in an extremely positive QBO phase expected to last through winter.  This is not considered a major factor but it does make it less likely we see prolonged blocking through winter.  

Solar:  We are still near a solar maximum expected to peak in 2025.  However, I have found using solar to predict blocking to be dangerous as the correlation is fairly low and hard to predict if the episode of blocking correlated with the cycle will happen in any given year.  Example:  The most reliable trend I've been able to indicate is that we almost always get a year of extreme blocking near a solar minimum.  It's happened the last 5 cycles.   1977, 1987, 1996, 2010, 2021.  But sometimes the blocking comes right at the minimum, sometimes it was the year after the minimum.  So even with this most reliable trend, its hard to use it as a reliable predictive tool as at best you have a 50/50 chance which doesn't help improve a seasonal forecasts chance of accuracy much.  For this reason I did not weight solar this year in my analog rankings.  I found when scoring my methods historically on what I would have predicted based on available data in the fall, I was better off not using solar at all.  

Methodological Changes

I won't go into all the details but I've significantly increased the weighting of my PDO in analog selections.  But more importantly, when scoring the PDO, our overall longer term phase of the PDO is as important as the most recent raw numerical number.  Not all -1.7 PDO's are the same...there is a big difference between a short term negative drop in the PDO during a more neutral or even positive long term cycle...v the same number during a strongly negative PDO cycle.  I've adjusted my scoring of the PDO to account for this.  

Forecast:

We all know the negatives...the combination of enso and pdo that we are in is not good and we know this because we've been suffering through it for most of the last 8 years.  So...probably the only question left is does this pattern flip that's just taken place change what the data is telling us about the coming winter....

Yes, but not as much as most of us would probably like.  When I factored in some of the most recent pattern analysis into the equations, it did change the analog rankings some.  And yes, some of the absolute worst analogs did drop out, but they weren't replaced by blockbuster winters by any means.  There just aren't many of those in the weak cold enso, -PDO category.  That combo is simply the absolute worst one for snow...and even if we pick out the best analog options from that combo...they just aren't very good.  

The one exception would be 2014, but its pretty far down on my analogs, that said with the new scoring, it was the next one tied with 2 others that just barely didn't make the cut.  The fall pattern progression similarities helped bump it up some, but it still missed the cut.  And the others that missed with it weren't nearly as good.  Even if I included all 17 seasons I looked at when I started identifying analogs before eliminating many of them...NONE of them came even close to 2014 in terms of snow.  It's simply an extreme outlier.  Someday it will happen again but my guess is not in my lifetime.  But who knows...

What years did make the cut.  Below is my analog rankings along with the weighting of the 4 point score for each category. 

So what about this upcoming great looking pattern...guess what all those analogs featured something similar...

October was mostly a torch, November was a transitions...and December was very cold

Fall-Early Winter Pattern for the top 6 analogs

October Mean Pattern 

But look what it lead to in December....and around November 20 was the transition period also! 

Dec Temps

Dec H5 Mean 

What about the rest of the winter.... well the pattern itself wasn't awful...on the whole they were near normal temps and all featured some cold periods.  And they did all feature SOME snow...but the results in every single one was worse that you would expect given the way they started in December and the generally decent to good pattern mean for the winter as a whole.  

Winter H5

Winter Temps 

The general tone of all those winters was generally near normal temps overall but with several periods of true cold, but dry and not a lot of big snowstorms.  Several of those years featured more snow south and north of us.  Southern sliders missing south and northern stream storms missing north was a feature of all these years.  

Snowfall Forecast

The raw predicted snowfall based on the analogs and weighting and my realistic snowfall range expectations

DCA: 7.1"  (5-10")

BWI: 8.3"  (6-12")

IAD: 9.6" (7-13")

RIC: 6.3" (4-9")

Manchester: 24.7"  (20-30")

Northern Shenandoah Valley (10-18")

NYC: 30.5"

You can see that snowfall increases dramatically just north of us.  While the changes in the calculations based on the current pattern did help us...it bumped up our expected snowfall a couple inches across the board and eliminated all the complete no snow dud years...it really helped places just north of us.  It significantly increased the snowfall projections for Philly to Boston for example.  We are just too far south to typically cash in on the northern stream dominant patterns even in the "colder" variety of nina, which I believe this will be.  

Temperature Forecast

Overall near normal temps

December Below Avg Temps

January: Near normal temps

Feb: Above normal Temps 

March: Below Normal Temps